Table 2 The estimated average treatment effect of a flames photo on homeowners’ information-seeking behavior in the field experiment.

From: Wildfire imagery reduces risk information-seeking among homeowners as property wildfire risk increases

 

Average treatment effect (ATE)

Conditional ATE: risk score

Conditional ATE excluding outliers

Model

1

2

3

4

5

6

Flames Photo

−0.009

−0.007

0.084*

0.091**

0.096**

0.090**

(0.010)

(0.010)

(0.033)

(0.033)

(0.036)

(0.034)

Risk Score/100

0.038**

0.037**

0.048**

0.048**

0.051**

0.048**

(0.004)

(0.004)

(0.005)

(0.005)

(0.006)

(0.005)

Flames photo × risk score/100

—

—

−0.021**

−0.022**

−0.023**

−0.022**

  

(0.008)

(0.008)

(0.008)

(0.008)

Ownership Characteristics

No

Yes

No

Yes

Yes

Yes

Parcel Characteristics

No

Yes

No

Yes

Yes

Yes

Observations

5785

5701

5785

5701

5571

5618

  1. Note: The estimated effect is the difference in webpage visitation (proportion) between the flames photo and the status quo photo. Linear probability estimation with robust standard errors. The estimated moderating effect of Risk Score is for a 100-point change in a parcel’s wildfire risk score. Models 1 and 2 show the main effects of the flames photo without and with household-level covariates. Models 3 and 4 show the effects of the flames photo conditional on parcel-level risk score. Results in the manuscript are reported from Model 4. Additional specifications exclude possible outliers of risk score (very low and very high scores) using the Interquartile Range criterion (Model 5), or observations greater than the 97.5th percentile or less than the 2.5th percentile (Model 6). Ownership characteristics include part-time ownership and owning multiple properties. Parcel characteristics include wildfire risk score, year built, acreage, and value. Standard errors in parentheses. — omitted from the regression. *Statistical significance at the 5% level; **statistical significance at the 1% level.
Back to article page